Research News
NYU Tandon researchers adapt cybersecurity tool to monitor health through smartphone traffic
Borrowing a cybersecurity technique originally designed to catch malware, researchers at NYU Tandon School of Engineering have developed a new tool for monitoring human health without invasive wearables or unreliable self-reporting.
The method, dubbed RouterSense, passively analyzes encrypted network traffic on a person's smartphone or other digital device, tracking their digital behaviors to potentially shed light on conditions ranging from mental health struggles in young adults to early signs of Alzheimer's disease.
"Traffic patterns serve as proxies for digital biomarkers," explained Danny Huang, the project’s senior researcher. "Screen time indicates sleep patterns, texting frequency reflects social interaction, and app usage reveals productivity rhythms, for example."
Huang is an NYU Tandon assistant professor with appointments in the Electrical and Computer Engineering Department (ECE) and in the Computer Science and Engineering Department. He is also on the faculties of the NYU Center for Cybersecurity, the Center for Urban Science + Progress and the Center for Advanced Technology in Telecommunications.
Because the system only captures metadata — information about digital activity such as which apps are contacted rather than the activity itself — it never sees any actual content, keeping the person's messages, videos, and other online activity private. The approach works across a diverse set of devices: phones, tablets, PCs, whether they run on Apple, Android, or Windows systems.
The analysis RouterSense employs has long been used in cybersecurity to spot malware by detecting unusual communication patterns, such as a device suddenly sending data to an unknown server or making suspicious connection requests. RouterSense flips this, tracking an individual's normal smartphone patterns instead of flagging threats.
"For the past fifteen years, I've been using network traffic analysis to understand how cybercriminals behave," said Huang. "My prior work has demonstrated that network traffic analysis could reveal behavioral patterns at scale while protecting privacy, lessons we're now applying to healthcare."
The research, recently released as a preprint in the Journal of Medical Internet Research, promises notable benefits over current standard health monitoring. Self reports are often riddled with recall bias. Intrusive sensors and smartphone apps can drain batteries and make people acutely aware they're being monitored. RouterSense, by contrast, is low-cost, unobtrusive, and scalable to large populations.
In a proof-of-concept study, 38 NYU students installed a VPN app on their smartphones for two weeks, which routed all their internet traffic through a research server in an NYU lab for analysis. Of the 29 participants who contributed valid data, 27 remained active for more than five days, contributing an average of over 300 hours of monitored network traffic.
The system successfully captured daily activity rhythms and lifestyle patterns, including gaming habits and late-night food delivery use, demonstrating the feasibility of passive monitoring. Participants reported the system was easy to use with many noting they forgot they were being monitored.
"Our approach allows us to increase the bandwidth of patient monitoring for months to years," said lead author Rameen Mahmood, an NYU Tandon ECE Ph.D. candidate. "Network traffic analysis lets us do passive monitoring 24-7 over an extended period of time, which hasn't really been done before."
This visualization showcases data from two participants in the researchers' 14-day pilot study with 27 NYU students. Each ring represents a full day of passively monitored internet activity from their mobile phones. The dots illustrate 10-minute intervals of internet usage, with red indicating high activity and blue indicating low activity. Notice the clear difference in patterns: Participant A demonstrates a consistent daily routine, including a regular sleep schedule (the prominent blue region at night), while Participant B's activity is more varied. Click to view larger image.
The NYU Tandon study's success in demonstrating feasibility and acceptability has paved the way for clinical applications.
The researchers are now recruiting individuals with pre-Alzheimer's conditions for a 30-day monitoring study comparing their network traffic patterns with healthy controls, the crucial next step in determining whether RouterSense can detect early signs of cognitive decline.
Other areas the researchers plan to explore include digital behaviors related to mindless eating and brain development in younger populations.
In addition to Huang, Mahmood and Kaye, the paper’s authors are Donghan Hu and Annabelle David (NYU Tandon), Zachary Beattie (Oregon Health & Science University), Nabil Alshurafa (Northwestern University), Lou Haux (Max Planck Institute for Human Development), Josiah Hester (Georgia Institute of Technology), Andrew Kiselica (University of Georgia), Shinan Liu (University of Hong Kong), and Chenxi Qiu and Chao-Yi Wu (Harvard Medical School).
NYU Tandon builds FAIR (Findable, Accessible, Interoperable, and Reusable) data infrastructure for urban monitoring
Urban infrastructure is constantly changing through weathering, deterioration, and repairs, yet the technology to track exactly what's happening to buildings and roads has remained frustratingly limited.
Now, a NYU Tandon-led research team with collaborators at the University of Delaware have received a grant from the National Science Foundation to build the missing data infrastructure that could help cities better monitor their built environments and identify problems earlier.
The project is called HS-SPECTRA (Hyperspectral Standardizing and Sharing Possibilities for Urban Conditions through Toolkits, Resources, and Archiving).
Hyperspectral imaging measures how light interacts with materials by recording how much light is reflected, absorbed, or transmitted at each wavelength across a broad spectrum. This process produces a unique signature for every material at its current condition. As such, the technique can reveal hidden moisture damage, structural repairs, and material deterioration invisible to the naked eye. But to use this technology effectively, researchers need reference databases showing what normal and damaged materials look like under real-world conditions.
Existing spectral libraries containing urban material signatures are extremely scarce and have no established standard for rigorous metadata, according to a review by Jessica Salcido and Debra Laefer published in the Photogrammetric Record. Furthermore, most existing spectral libraries capture materials once, under idealized laboratory conditions, making those libraries incomplete and hard to use.
"We're creating something entirely new," said Laefer, the project's principal investigator and NYU Tandon professor of Civil and Urban Engineering and a faculty member of the Center for Urban Science + Progress (CUSP). "A living dataset that tracks the same materials repeatedly over time and under varying environmental conditions and temperatures."
The practical applications extend across multiple domains. Building inspectors could use the technology to identify structural repairs or detect moisture damage invisible to the naked eye. Urban planners might track how infrastructure weathers over time. Heritage conservators could monitor historic buildings for early signs of deterioration.
“You can look at a brick facade with your eyes and maybe notice some discoloration. But hyperspectral imaging can distinguish between original materials and patches, identify salt intrusion, and detect moisture retention,” said Salcido, an NYU Physics Ph.D candidate and a researcher on the project. “That information is critical for maintenance planning but impossible to see with conventional photography."
HS-SPECTRA will employ rooftop-mounted sensing platforms in Brooklyn at CUSP's Urban Observatory and at the University of Delaware. The one at NYU will provide persistent imaging of the Manhattan skyline, as the observatory has the capability to capture visible, infrared, and hyperspectral data at distances of one to four kilometers. The facility at the University of Delaware is led by Prof. Gregory Dobler, who is the co-principal investigator on this NSF grant.
The team will augment this existing infrastructure with additional weather and atmospheric sensors to fully contextualize how environmental conditions affect spectral measurements. Over the three-year, $600,000 project, they expect to generate approximately 102,000 individual spectra annually from multiple carefully selected regions of interest, capturing hourly data between 6 AM and 8 PM across all seasons.
"We're creating open-source tools and standards that will enable seamless conversion of our reference data to the research instruments owned and operated by other scientists and engineers around the world," Laefer explained, addressing that what sets HS-SPECTRA apart is making the data actually usable.
The project will do this by developing standardized metadata frameworks, processing pipelines, and automated archival workflows. All data will be released through established public repositories, SPECCHIO and EcoSIS, with scheduled releases linked by unified campaign identifiers. The team is also creating resampling software that will enable researchers to transform the spectral library to match specifications of various sensors, dramatically expanding the data's utility.
The project includes community engagement efforts, with case studies, training materials, and demonstrations targeted at the structural engineering, architecture, and historic preservation communities through professional conferences, webinars, and direct outreach to city agencies like NYC's Department of Buildings. By automating the data processing and archival pipeline, the team aims to ensure the urban spectral monitoring can continue well beyond the grant period, creating a lasting resource for researchers worldwide studying how cities evolve at the material level.
NSF's Office of Advanced Cyberinfrastructure and the Directorate for Engineering jointly support the project.
NYU Tandon researchers are developing AI-powered exoskeletons to enhance human mobility for everyone
Approximately 31 million Americans experience mobility disabilities, with nearly 27% of older adults facing significant difficulty walking or climbing stairs. Yet current robotics solutions remain largely confined to research laboratories.
The fundamental problem, as Hao Su at NYU Tandon School of Engineering explains, is that existing exoskeletons "don't really understand humans’ intention." Most rehabilitation exoskeletons require 30 minutes or more to calibrate for each user and are uncomfortably bulky and heavy at around or more than 30 pounds, making people reluctant to wear them.
Su is leading a multi-university research team that is pursuing a solution through a recently awarded $3.6 million collaborative NSF Growing Convergence Research grant.
Building on a 2024 study Su and his colleagues published in Nature that proved exoskeletons can leverage deep reinforcement learning to train control policies entirely in computer simulation, the new project aims to take this technology from healthy-subject laboratory demonstrations into real-world community use for healthy people, older adults, and those with disabilities.
Ultimately, the team’s goal is to create affordable, lightweight exoskeletons that are intelligent enough to adapt to the needs of older adults and stroke survivors without requiring lengthy setup, while also being comfortable enough for daily use in their communities.
"This is truly an interdisciplinary endeavor. Our work involves mechatronics design, electrical engineering, computer science, rehabilitation medicine, and gerontology," explains Su, an associate professor of Biomedical Engineering.
The team's approach, which Su refers to as "physical AI," represents a fundamental shift from traditional methods. "Standard approaches require collecting extensive training data from 20 to 40 people for each study. In our Nature paper (see video below), we showed that our system requires no physical data collection for training. Instead, we utilized publicly available data and trained our system to accommodate a wide range of users. In this new project, we will use wearable sensors or cameras in smartphones to capture human movements to personalize control policies for each individual."
The approach uses a highly detailed virtual human body that simulates how hundreds of individual muscles and over 50 different joint movements work together. Using this virtual body, the team trains three AI systems that each handle different tasks: one learns natural human movement patterns, another determines which muscles should activate during movements, and a third decides what assistance the exoskeleton should provide in real-time.
The efficiency of this approach is remarkable. "The AI system that controls the exoskeleton only needs to train once for a few hours in the simulation. During this learning process, it gradually learned to generate effective assistance for walking, running, and stair climbing activities. In this new project, we will expand knowledge of assistive controllers to pathological gaits and more activities," Su explains.
The Nature study showed that people wearing the exoskeleton used 24.3% less energy during walking, 13.1% less during running, and 15.4% less during stair climbing. The mechatronics innovation of their robots is equally significant. Su's team has systematically reduced complexity and weight.
"We eliminated, for example, an expensive torque sensor by developing a new algorithm that estimates the same measurements using software instead of hardware. This makes the device lighter and much cheaper," Su explains. The team has built approximately 30 iterations of the device, each progressively lighter and more comfortable. The team’s version weighs about 6.6 pounds, as published in IEEE/ASME Transactions on Mechatronics 2025, compared to 30-pound tethered systems that confine users to laboratory treadmills.
The project’s vision extends beyond medical applications. "Our goal is what we call exoskeletons for everyone and everywhere,” said Su. “It's possible exoskeletons could be used, for example, for recreational activities like hiking. Or simply to make it easier for anyone to walk farther or longer distances in their daily lives."
The research team’s co-principal investigators are Xianlian Alex Zhou at New Jersey Institute of Technology (NJIT), Pamela Cacchione and Michelle Johnson at the University of Pennsylvania, Xiaoyue Ni at Duke University, and Bolei Zhou at UCLA.
Yan Y, Huang JS, Zhu J, Hou Z, Gao W, Lopez-Sanchez I, Srinivasan N, Srihari A, Su H. Compact and Foldable Hip Exoskeleton With High Torque Density Actuator for Walking and Stair-Climbing Assistance in Young and Elderly Adults. IEEE/ASME Transactions on Mechatronics. 2025 Jul 8.
Researchers demonstrates substrate design principles for scalable superconducting quantum materials
Silicides — alloys of silicon and metals long used in microelectronics — are now being explored again for quantum hardware. But their use faces a critical challenge: achieving phase purity, since some silicide phases are superconducting while others are not.
The study, published in Applied Physics Letters by NYU Tandon School of Engineering and Brookhaven National Laboratory, shows how substrate choice influences phase formation and interfacial stability in superconducting vanadium silicide films, providing design guidelines for improving material quality.
The team, led by NYU Tandon professor Davood Shahrjerdi, focused on vanadium silicide, a material that becomes superconducting (able to conduct electricity without resistance) when cooled below its transition temperature of 10 Kelvin, or about -263°C. Its relatively high superconducting transition temperature makes it attractive for quantum devices that operate above conventional millikelvin temperatures.
Researchers engineered crystalline hafnium oxide substrates and compared them with standard silicon dioxide under identical processing conditions. Hafnium oxide offered greater chemical stability and suppressed unwanted secondary phases, though it degraded under the highest processing temperatures.
"Achieving phase-pure superconducting films requires careful attention to the substrate-film interface," said Shahrjerdi. "Our findings show that substrate design is an integral aspect of the synthesis process.”
The chemical stability of hafnium oxide proved crucial for maintaining film quality during processing. Most intriguingly, atomic-resolution imaging suggested that the crystalline structure of hafnium oxide may influence the orientation and phase selection of overlying silicide grains, pointing to possible templating effects that could enable selective phase nucleation.
The research provides fundamental insights that extend beyond vanadium silicides to other superconducting silicide systems. The principles identified — chemical inertness, thermal stability, and structural ordering — offer design guidelines for next-generation quantum device substrates.
"These findings complement our recent work on physical patterning techniques," noted Shahrjerdi. "Together, they expand the design space for quantum hardware."
In addition to Shahrjerdi, the paper’s authors are Miguel Manzo-Perez, Moeid Jamalzadeh, and Iliya Shiravand (Ph.D. students at NYU Tandon); and Sooyeon Hwang, Kim Kisslinger, and Dmytro Nykypanchuk from the Center for Functional Nanomaterials at Brookhaven National Laboratory. The work was conducted in part at the NYU Nanofabrication Cleanroom (NYU Nanofab) with characterization support from Brookhaven National Laboratory.
Miguel Manzo-Perez, Moeid Jamalzadeh, Sooyeon Hwang, Iliya Shiravand, Kim Kisslinger, Dmytro Nykypanchuk, Davood Shahrjerdi; Substrate effects on phase formation and interfacial stability in superconducting vanadium silicide thin films. Appl. Phys. Lett. 22 September 2025; 127 (12): 122601. https://doi.org/10.1063/5.0291576
New NYU Tandon-led project will accelerate privacy-preserving computing
Today's most advanced cryptographic computing technologies — which enable privacy-preserving computation — are trapped in research labs by one critical barrier: they're thousands of times too slow for everyday use.
NYU Tandon, helming a research team that includes Stanford University and the City University of New York, just received funding from a $3.8 million grant from the National Science Foundation to build the missing infrastructure that could make those technologies practical, via a new design platform and library that allows researchers to develop and share chip designs.
The problem is stark. Running a simple AI model on encrypted data takes over 10 minutes instead of milliseconds, a four order of magnitude performance gap that impedes many real-world use cases.
Current approaches to speeding up cryptographic computing have hit a wall, however. "The normal tricks that we have to get over this performance bottleneck won’t scale much further, so we have to do something different," said Brandon Reagen, the project's lead investigator. Reagen is an NYU Tandon assistant professor with appointments in the Electrical and Computer Engineering (ECE) Department and in the Computer Science and Engineering (CSE) Department. He is also on the faculty of NYU's Center for Advanced Technology in Telecommunications (CATT) and the NYU Center for Cybersecurity (CCS).
The team's solution is a new platform called "Cryptolets.”
Currently, researchers working on privacy chips must build everything from scratch. Cryptolets will provide three things: a library where researchers can share and access pre-built, optimized hardware designs for privacy computing; tools that allow multiple smaller chips to work together as one powerful system; and automated testing to ensure contributed designs work correctly and securely.
This chiplet approach — using multiple small, specialized chips working together — is a departure from traditional single, monolithic chip optimization, potentially breaking through performance barriers.
For Reagen, this project represents the next stage of his research approach. "For years, most of our academic research has been working in simulation and modeling," he said. "I want to pivot to building. I’d like to see real-world encrypted data run through machine learning workloads in the cloud without the cloud ever seeing your data. You could, for example, prove you are who you say you are without actually revealing your driver's license, social security number, or birth certificate."
What sets this project apart is its community-building approach. The researchers are creating competitions where students and other researchers use Cryptolets to compete in designing the best chip components. The project plans to organize annual challenges at major cybersecurity and computer architecture conferences. The first workshop will take place in October 2025 at MICRO 2025, which focuses on hardware for zero-knowledge proofs.
"We want to build a community, too, so everyone's not working in their own silos," Reagen said. The project will support fabrication opportunities for competition winners, with plans to assist tapeouts of smaller designs initially and larger full-system tapeouts in the later phases, helping participants who lack chip fabrication resources at their home institutions
"With Cryptolets, we are not just funding a new hardware platform—we are enabling a community-wide leap in how privacy-preserving computation can move from theory to practice,” said Deep Medhi, program director in the Computer & Information Sciences & Engineering Directorate at the U.S. National Science Foundation. “By lowering barriers for researchers and students to design, share and test cryptographic chips, this project aligns with NSF’s mission to advance secure, trustworthy and accessible technologies that benefit society at large."
If the project succeeds, it could enable a future where strong digital privacy isn't just theoretically possible, but practically deployable at scale, from protecting personal health data to securing financial transactions to enabling private AI assistants that never see people's actual queries.
Along with Reagen, the team is led by NYU Tandon co-investigators Ramesh Karri, ECE Professor and Department Chair, and faculty member of CATT and CCS; Siddharth Garg, Professor in ECE and faculty member of NYU WIRELESS and CCS; Austin Rovinski, Assistant Professor in ECE; The City College of New York’s Rosario Gennaro and Tushar Jois; and Stanford's Thierry Tambe and Caroline Trippel, with Warren Savage serving as project manager. The team also includes industry advisors from companies working on cryptographic technologies.
NYU Tandon team help develop bio-inspired robotics for disaster response and construction, in new NSF-funded project
The United States recorded 28 natural disasters causing at least $1 billion in damages each in 2023, the highest number in the nation's history. Now researchers at NYU Tandon are helping develop a robotic system that could significantly reduce disaster recovery times while improving efficiency for contractors working in confined spaces.
Along with colleagues from New Jersey Institute of Technology, who led the project, and a researcher from The University of Scranton, the Tandon team led by Maurizio Porfiri and Semiha Ergan is part of a three-year, $5 million U.S. National Science Foundation (NSF)-funded project to create the Kastor robotic system. The funding comes from the NSF Directorate for Technology, Innovation and Partnerships, which supports research that brings together multiple disciplines and sectors to solve complex societal and operational challenges.
This Phase 2 award follows a previous $650,000 Phase 1 grant that developed a prototype robot and algorithms.
The Kastor robotic system uses swarms of self-assembling robots to transport equipment and clear debris in disaster zones, addressing a persistent challenge in disaster response: much of the workforce effort goes toward moving supplies and removing debris rather than critical tasks like searching for survivors.
The technology takes its design cues from fire ants and slime molds. Fire ants can link their bodies to form bridges over difficult terrain, while slime molds create efficient transport networks across varied surfaces. The Kastor system applies these biological strategies to create networks of flat metal robotic tiles that can autonomously reconfigure themselves as conditions change.
The tiles move themselves into position and use wheels and treads to transport pallets across disaster sites without human intervention. Algorithms developed by the research team guide their assembly and movement patterns.
Porfiri — who directs NYU's Center for Urban Science + Progress (CUSP) and is Institute Professor in the departments of Mechanical and Aerospace Engineering, Biomedical Engineering, and Civil and Urban Engineering (CUE) — brings expertise in urban science and virtual reality to the project. His role focuses on ensuring the technology integrates with existing disaster response workflows in urban environments.
Ergan — an associate professor in CUE, and on the faculty of CUSP, Institute of Design and Construction (IDC) Innovation Hub, and C2SMARTER transportation center — is leading virtual and on-site pilot studies that will test the system in realistic construction and recovery scenarios.
"Each community faces different challenges when disasters strike, and current response methods often require inefficient manual labor for debris removal and supply transport," Porfiri said. The project team has consulted with police officers, emergency responders, contractors and construction companies to understand operational requirements.
"We want to bring the high-tech automation of distribution facilities and smart warehouses to messy, unstructured outdoor environments," said Petras Swissler, an assistant professor of mechanical and industrial engineering at NJIT and the project's principal investigator.
Beyond disaster response, the researchers found the same challenges exist in construction projects, where efficiency improvements have lagged behind other industries.
"This technology will also assist at construction sites where space is tight and the ability to navigate in multiple directions while carrying dirt and construction materials is limited," Ergan said.
The project will develop a production-ready robotic system, create interfaces for operators to control the robot swarms, and conduct pilot studies in both disaster response and construction settings. Along with Porfiri and Ergan, the other co-principal investigators are Simon Garnier, a biology professor at NJIT, and Jason Graham, a mathematics professor at The University of Scranton.
An eco-friendly way to see in the dark
Manufacturers of infrared cameras face a growing problem: the toxic heavy metals in today's infrared detectors are increasingly banned under environmental regulations, forcing companies to choose between performance and compliance.
This regulatory pressure is slowing the broader adoption of infrared detectors across civilian applications, just as demand in fields like autonomous vehicles, medical imaging and national security is accelerating.
In a paper published in ACS Applied Materials & Interfaces, researchers at NYU Tandon School of Engineering reveal a potential solution that uses environmentally friendly quantum dots to detect infrared light without relying on mercury, lead, or other restricted materials.
The researchers use colloidal quantum dots which upends the age-old, expensive, and tedious processing of infrared detectors. Traditional devices are fabricated through slow, ultra-precise methods that place atoms almost one by one across the pixels of a detector — much like assembling a puzzle piece by piece under a microscope.
Colloidal quantum dots are instead synthesized entirely in solution, more like brewing ink, and can be deposited using scalable coating techniques similar to those used in roll-to-roll manufacturing for packaging or newspapers. This shift from painstaking assembly to solution-based processing dramatically reduces manufacturing costs and opens the door to widespread commercial applications.
"The industry is facing a perfect storm where environmental regulations are tightening just as demand for infrared imaging is exploding," said Ayaskanta Sahu, associate professor in the Department of Chemical and Biomolecular Engineering (CBE) at NYU Tandon and the study's senior author. "This creates real bottlenecks for companies trying to scale up production of thermal imaging systems."
Another challenge the researchers addressed was making the quantum dot ink conductive enough to relay signals from incoming light. They achieved this using a technique called solution-phase ligand exchange, which tailors the quantum dot surface chemistry to enhance performance in electronic devices. Unlike traditional fabrication methods that often leave cracked or uneven films, this solution-based process yields smooth, uniform coatings in a single step — ideal for scalable manufacturing.
The resulting devices show remarkable performance: they respond to infrared light on the microsecond timescale — for comparison, the human eye blinks at speeds hundreds of times slower — and they can detect signals as faint as a nanowatt of light.
"What excites me is that we can take a material long considered too difficult for real devices and engineer it to be more competitive," said graduate researcher Shlok J. Paul, lead author on the study. "With more time this material has the potential to shine deeper in the infrared spectrum where few materials exist for such tasks."
This work adds to earlier research from the same lead researchers that developed new transparent electrodes using silver nanowires. Those electrodes remain highly transparent to infrared light while efficiently collecting electrical signals, addressing one component of the infrared camera system.
Combined with their earlier transparent electrode work, these developments address both major components of infrared imaging systems. The quantum dots provide environmentally compliant sensing capability, while the transparent electrodes handle signal collection and processing.
This combination addresses challenges in large-area infrared imaging arrays, which require high-performance detection across wide areas and signal readout from millions of individual detector pixels. The transparent electrodes allow light to reach the quantum dot detectors while providing electrical pathways for signal extraction.
"Every infrared camera in a Tesla or smartphone needs detectors that meet environmental standards while remaining cost-effective," Sahu said. "Our approach could help make these technologies much more accessible."
The performance still falls short of the best heavy-metal-based detectors in some measurements. However, the researchers expect continued advances in quantum dot synthesis and device engineering could reduce this gap.
In addition to Sahu and Paul, the paper's authors are Letian Li, Zheng Li, Thomas Kywe, and Ana Vataj, all from NYU Tandon CBE. The work was supported by the Office of Naval Research and the Defense Advanced Research Projects Agency.
Paul, S. J., Li, L., Li, Z., Kywe, T., Vataj, A., & Sahu, A. (2025). Heavy Metal Free Ag2Se Quantum Dot Inks for Near to Short-Wave Infrared Detection. ACS Applied Materials & Interfaces. doi:10.1021/acsami.5c12011
NYU Tandon researchers launch interactive 3D flood map to help New Yorkers visualize climate risks
When Hurricane Sandy devastated New York City in 2012, it became clear that communicating flood risk through traditional probability maps wasn't enough.
Now, researchers at NYU Tandon School of Engineering have created GeoFlood Studio, an interactive 3D flood visualization platform that lets users see exactly how water would rise around their neighborhood during major storms.
"Talking about probability and technical terms in risk assessment often isn't digestible," said Yuki Miura, Assistant Professor in the Department of Mechanical and Aerospace Engineering and the Center for Urban Science + Progress, who leads the project. "We need realistic tools that people can understand."
GeoFlood Studio, developed over the past two months by Miura's Climate, Energy, and Risk Analytics (CERA) Lab, represents a significant advance over existing flood visualization tools. “While other 3D flood models exist, they typically offer limited scenarios, slow loading times, and basic flood-zone style information that shows only which areas might be underwater, without the level of detail on depth, velocity, and human vulnerability that our platform provides,” said Miura.
With GeoFlood Studio, users can explore scenarios based on Hurricane Sandy (coastal flooding) and Hurricane Ida (rainfall-driven flooding), each combined with projected sea level rise for 2050, 2080, and 2100. The tool allows examination of compound flooding scenarios where both storm types occur simultaneously.
The platform's interactive features include adjustable human silhouettes to show precisely how deep flood waters would reach at the user's height. Users can toggle velocity overlays to see how fast water is projected to move, or vulnerability overlays that show color-coded safety zones ranging from areas safe for driving to zones where adult life is in danger.
Users can also enable or disable the East Side Coastal Resiliency seawall to instantly see how this infrastructure protects against flooding.
"If flood depths are about two feet, you might think that’s not so dangerous. But if the velocity is high, the risk can be much greater than you expect. Conversely, if the velocity is low, you may still face challenges but have a better chance of safety," Miura explained.The map loads immediately, allowing users to click anywhere for instant data about flood depth, water velocity, and vulnerability (danger) levels. Users can select preset viewing locations across Lower Manhattan or navigate the 3D map freely.
Currently focused on Lower Manhattan, the platform is expanding rapidly. Within weeks, GeoFlood Studio will include evacuation routing capabilities, showing the safest path to emergency shelters while avoiding dangerous flood zones. The team plans to expand coverage to all of New York City within months and enable other researchers to upload their own flood scenarios by year's end.
Applications extend beyond academic research. Emergency management agencies can plan evacuation routes, insurance companies can assess property risks more accurately than traditional FEMA flood zone maps, and real estate developers can evaluate how proposed seawalls might protect properties.
"Asset management firms and insurance companies can use this to see exactly which areas will be underwater and how deep, going beyond the broad categories provided by FEMA’s flood zone maps," Miura noted.
The project reflects Miura's broader research mission to "identify, measure, and manage risks" through tools that work for both technical experts and community members. GeoFlood Studio is part of her larger body of work modeling urban flood risks, which includes developing rapid flood estimation tools that deliver real-time flood forecasts within seconds.
The team has gathered feedback from New York City agencies including the Department of Environmental Protection and the Mayor's Office, as well as local high school students and artists, ensuring the platform serves diverse users.
As climate change intensifies flood risks worldwide, GeoFlood Studio offers a model for making complex climate science accessible and actionable. By letting people see themselves in flood scenarios rather than reading statistics, the platform transforms abstract risk into a tangible, personal understanding.
NYU will showcase GeoFlood Studio during a public workshop on September 26, 2025 as part of NYU Climate Week 2025, inviting residents, practitioners, and policymakers to explore the tool and discuss how visual insights can drive climate adaptation planning.
New York City's medical specialist advantage may be an illusion, new NYU Tandon research shows
New York City offers nearly every type of medical specialist but provides fewer specialty healthcare providers per capita than smaller cities, according to a new study that challenges conventional assumptions about urban healthcare advantages and reveals a troubling paradox across America's largest metropolitan areas.
The research, published in Nature Cities, analyzed data from 1.4 million healthcare providers across 75 medical specialties in 898 metropolitan and micropolitan areas. The innovative approach combines urban scaling theory—which examines how city characteristics change with population size—with network science and economic geography to examine healthcare access in unprecedented detail.
Rather than treating healthcare as a single entity, the researchers examined each medical specialty separately, revealing that 88% exhibit what they call "sublinear scaling," meaning larger cities have proportionally fewer specialists per resident than smaller ones.
"We're discovering that the healthcare advantages of living in big cities may be an illusion when it comes to specialized care," explains lead researcher Maurizio Porfiri. "We all assume residents of large metropolitan areas have better access to healthcare than residents of smaller cities, but this is really true only for primary care services. Our findings suggest this assumption breaks down completely for medical specialists. A small city may not offer all the specialties of large cities, but in what it offers it may outperform them.”
Porfiri is an NYU Tandon Institute Professor with appointments in the Departments of Mechanical and Aerospace Engineering (MAE), Biomedical Engineering (BME), Civil and Urban Engineering (CUE), and Technology Management and Innovation (TMI). He also serves as Director of the NYU Center for Urban Science + Progress (CUSP).
The study represents the latest application of Porfiri's urban scaling methodology, which he has previously used to analyze gun violence patterns and the relationship between city living, ADHD and obesity. His research uses Scale-Adjusted Metropolitan Indicators (SAMIs) to control for population differences and reveal how cities deviate from expected patterns.
The study found that while cities like New York and Chicago offer nearly all examined specialties (NYC has 74 — missing only anesthesiology assistants — and Chicago has all 75), residents may face longer wait times and specialists higher patient loading.
In contrast, smaller cities may lack certain specialties entirely—73 of the 75 specialties showed significant associations between availability and population size—but those that exist serve fewer patients per provider. For example, Marshfield, Wisconsin provides 16.8 specialists per 1,000 residents compared to New York's 4.7 per 1,000.
Among the most underrepresented specialties in large cities per capita are addiction medicine, preventive medicine, osteopathic manipulative medicine, and micrographic dermatologic surgery.
Addiction medicine shows the starkest disparity, with large cities providing dramatically fewer specialists per resident than smaller areas. These fields showed the strongest sublinear scaling, meaning residents of major metropolitan areas have significantly fewer of these specialists available relative to their population size compared to smaller cities.
The research identifies two mechanisms driving this paradox: higher patient loads overwhelming specialists in large cities, and economic clustering that concentrates medical expertise in dense hospital networks, creating geographic inequalities.
“The findings have serious implications as the U.S. population ages. The study found sublinear scaling in geriatric specialties like urology and gerontology, suggesting major metropolitan areas may be unprepared for growing elderly populations,” said Tian Gan, a NYU Tandon mechanical engineering PhD student in the urban science track, and the paper’s lead author.
Geographic patterns reveal stark regional disparities. The highest specialist concentrations cluster in the Midwest—Minnesota alone claims two of the top five cities—while all five cities with the lowest access are in the South.
Not all specialties follow this pattern. Several key specialties—including anesthesiology, internal medicine, and clinical psychology—actually have more providers per capita in large cities, reflecting higher urban demand for these services.
The research provides a framework for understanding healthcare distribution that moves beyond the traditional urban-rural dichotomy. Rather than viewing cities as uniformly advantaged, policymakers must consider the complex interplay between diversity and provision of medical services.
Along with Porfiri and Gan, the paper's additional author is Tanisha Dighe, NYU Tandon MS student in applied urban science and information. The study was supported by National Science Foundation grants.
APPENDIX: Medical Specialist Availability by City
CITIES WITH THE MOST MEDICAL SPECIALISTS (Cities offering all specialty types)
- Chicago-Naperville-Elgin, IL-IN: 75 specialties
- Houstone-Pasadena-The Woodlands, TX: 75 specialties
- Atlanta-Sandy Springs-Roswell, GA: 75 specialties
- Washington-Arlington-Alexandria, DC-VA-MD-WV: 75 specialties
- Miami-Fort Lauderdale-West Palm Beach, FL: 75 specialties
CITIES WITH THE FEWEST MEDICAL SPECIALISTS (Fewest specialty types available)
- Monroe, LA: 5 specialties
- Zapata, TX: 6 specialties
- Raymondville, TX: 6 specialties
- Synder, TX: 11 specialties
- Andrews, TX: 11 specialties
CITIES WITH THE HIGHEST CONCENTRATION OF SPECIALISTS OVERALL (All non-primary-care specialists combined per 1,000 residents)
- Rochester, Minnesota: 21.1 specialists (home to Mayo Clinic)
- Marshfield, Wisconsin: 16.8 specialists
- Sunbury, Pennsylvania: 16.3 specialists
- Easton, Maryland: 15.7 specialists
- Albert Lea, Minnesota: 15.4 specialists
CITIES WITH THE LOWEST CONCENTRATION OF SPECIALISTS OVERALL (Fewest specialists per 1,000 residents)
- Monroe, Louisiana: 0.1 specialists
- Virginia Beach-Norfolk, Virginia: 0.4 specialists
- Danville, Virginia: 0.8 specialists
- Rio Grande City-Roma, Texas: 1.0 specialists
- Bonham, Texas: 1.0 specialists
SPECIALTIES MOST UNDERREPRESENTED IN MAJOR METROS, 1M+ POPULATION
(Scaling exponents - how fast they grow with population growth )
- Addiction Medicine (0.305) - Most underrepresented
- Preventive Medicine (0.331)
- Osteopathic Manipulative Medicine (0.351)
- Micrographic Dermatologic Surgery (0.379)
- Maxillofacial Surgery (0.398)
- Marriage and Family Therapist (0.400)
- Nuclear Medicine (0.408)
- Advanced Heart Failure and Transplant Cardiology (0.446)
- Certified Clinical Nurse Specialist (0.457)
- Sleep Medicine (0.457)
SPECIALTIES MOST OVERREPRESENTED IN MAJOR METROS
(Scaling exponents - - how fast they grow with population growth)
- Anesthesiology (1.154) - Most overrepresented
- Internal Medicine (1.100)
- Physical Therapy (1.089)
- Clinical Psychology (1.069)
- Physician Assistant (1.057)
- Obstetrics/Gynecology (1.050)
- Neurology (1.039)
- Psychiatry (1.031)
- Gastroenterology (1.022)
NYC SPECIALIST COUNTS (74 out of 75 research specialties)
Missing only: Anesthesiology Assistant
Top 10:
- Nurse Practitioner: 8,977
- Internal Medicine: 8,194
- Physical Therapy: 7,515
- Physician Assistant: 6,224
- Clinical Social Worker: 4,842
- Anesthesiology: 3,637
- Family Practice: 3,259
- Diagnostic Radiology: 2,843
- Emergency Medicine: 2,545
- Psychiatry: 2,465
Notable underrepresented specialties (bottom 5):
- Maxillofacial Surgery: 40
- Micrographic Dermatologic Surgery: 25
- Preventive Medicine: 21
- Marriage and Family Therapist: 18
- Addiction Medicine: 16
Gan, T., Dighe, T. & Porfiri, M. Trade-off between diversity and provision of specialized healthcare in US cities. Nat Cities (2025).
David M. Truong’s Programmable Regenerative Immunity Lab is doing promising research that could open up new possibilities for Alzheimer’s patients
Assistant Professor of Biomedical Engineering David Truong and his research group are working to create programmable macrophage–neuron interfaces able to reduce the inflammation associated with Alzheimer’s disease and repair damaged neural connections.
Alzheimer’s involves complex immune–neural interactions that drive inflammation, synaptic loss, and cognitive decline, Truong explains, and current therapies target only isolated pathological features, failing to address the broader breakdown in neuroimmune signaling; he intends to address those shortcomings by engineering macrophage-specific circuits that secrete therapeutic payloads in response to small molecule inducers, enabling tunable neuroimmune repair.
The project builds upon Truong’s recent NIH-funded work, which uses engineered immune cells to target and remove amyloid plaques (one of the hallmarks of Alzheimer’s disease) when delivered through the bloodstream, where they cross the blood-brain barrier.
The immune cells used in his lab are “off-the-shelf,” meaning that they do not need to be taken from a patient but can instead be manufactured and prepared in advance; they are created from human induced pluripotent stem cells (iPSCs), a renewable source of cells that can be genetically modified in the lab.
His new project recently won “Early Stage” support from NYU’s Discovery Research Fund for Human Health, a program launched in late 2023 to aid faculty members in addressing significant medical challenges. (In its first full year, it provided support to teams researching interventions for diabetes prevention, improvements to targeted cancer therapies, and advanced nanofabrication techniques to enable the rapid detection of multiple pathogens, among other projects.)
Truong — whose laurels include a New Innovator Award from the National Institute of Allergy and Infectious Diseases, a Delil Nasser Award for Professional Development from the Genetics Society of America, and a National Institutes of Health Ruth L. Kirschstein National Research Service Award — envisions eventually creating an overall framework for regenerative neuroimmune therapies. “We may one day see even broader applications in neurodegeneration, immune dysfunction, and human brain repair,” he predicts.